Serum Metrnl is associated with the presence and severity of coronary artery disease

Abstract Meteorin‐like (Metrnl) is a novel adipokine that is highly expressed in white adipose tissue. Metrnl stimulates energy expenditure and improves glucose tolerance in rodents. However, whether Metrnl plays a role in coronary artery disease (CAD) remains to be elucidated. The present study aimed to investigate the association of serum Metrnl with CAD in Chinese patients. A total of 193 patients with CAD and 156 control subjects were enrolled in this study. Serum Metrnl concentration was measured by enzyme‐linked immunosorbent assay. Anthropometric phenotypes, fasting glucose, serum lipids, and inflammatory cytokines were measured. Serum Metrnl was lower in CAD patients when compared to those controls (132.41 vs 173.17 pg/mL, P < 0.001). Serum Metrnl was negatively correlated with metabolic parameters, including body mass index, total cholesterol, and low‐density lipoprotein cholesterol as well as inflammatory markers including high‐sensitivity C‐reactive protein, IL‐1β, and IL‐11 even after adjustment for potential confounding variables (P < 0.05). In multivariable logistic regression analyses, compared to those in the highest tertile of serum Metrnl levels, subjects in the lowest tertile had the highest risks for CAD (adjusted OR = 2.63, 95% CI = 1.46‐4.27, P = 0.001). After adjustment for potential confounding variables, serum Metrnl was also decreased as the number of stenosed vessels increased (P < 0.001). Furthermore, decreased Metrnl level was negatively correlated with the severity of CAD quantified by the Gensini score. This first case‐control study shows significant associations of serum Metrnl with the presence and severity of CAD, suggesting Metrnl might be a new promising therapeutic target for CAD.


| INTRODUCTION
Coronary artery disease (CAD) remains the leading cause of morbidity and mortality in the adult population worldwide. 1 Adipose tissue has been recognized as the largest endocrine and paracrine organ that produces hundreds of proteins contributing as enzymes, hormones, or growth factors in various pathophysiologic processes. 2 These proteins are known under the term "adipokines," for example, adiponectin, leptin, resistant, visfatin, retinol-binding protein 4, etc., which are involved in the pathophysiology of obesity and insulin resistance and therefore have indirect effects on atherosclerosis. 3,4 For instance, adiponectin is an adipose tissue-derived adipokine which has protective role in the initiation and progression of atherosclerosis through its anti-inflammatory and antiatherogenic effects. 5 Leptin exerts various atherogenic effects such as induction of endothelial dysfunction, stimulation of inflammatory reaction and oxidative stress, decrease in paraoxonase activity, platelet aggregation, etc. 6 Leptin-deficient and leptin receptor-deficient mice are protected from arterial thrombosis and neointimal hyperplasia in response to arterial wall injury. 7 Our previous studies indicated that visfatin, a pro-inflammatory adipokine expressed markedly in visceral fat, may promote atherosclerosis through stimulating vascular inflammation. 8,9 Recently, Meteorin-like (Metrnl), also known as Cometin and Subfatin, has been identified as a novel adipokine that is highly expressed in the subcutaneous white adipose tissue. 10 Rao et al 11

found that
Metrnl was induced in white adipose tissue during acute cold exposure and in muscle after acute bouts of exercise or overexpression of Peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α). Metrnl could promote beige/brown fat thermogenesis and improve glucose tolerance via activating IL-4/IL-13 signalling. 11 Another study showed that Metrnl was downregulated in white adipose tissue after caloric restriction but was dramatically upregulated during white adipocyte differentiation and in the white adipose tissue of diet-induced obese mice. 12 Li et al 13  Further experiments identified that Metrnl antagonized obesityinduced insulin resistance by improving adipose function, including adipocyte differentiation, metabolism activation, and inhibiting adipose inflammation through the PPARγ pathway. 13 A clinical study also showed that serum Metrnl level was decreased and negatively correlated with glucose level and insulin resistance in patients with diabetes. 14

| Coronary angiography
Two cardiologists who were unaware of the patients included in this study assessed the angiograms. CAD was defined as luminal diameter narrowing estimated visually at least 50% in any epicardial coronary artery, including the left main coronary artery, left anterior descending, left circumflex, or right coronary artery. CAD patients were divided into single-, double-, and triple-vessel disease subgroups according to the number of significantly stenosed vessels.
The severity of CAD was assessed with the Gensini score system based on the degree of luminal narrowing and its geographic importance. 18

| Laboratory measurements
Venous blood sample was collected from each subject. The whole blood was separated into serum and cellular fractions within 2 hours by centrifugation at 3000 g for 10 minutes. The supernatant (serum) was collected and further centrifuged at 10 000 g for 15 minutes to completely remove the cell debris. The obtained serum was stored at −80°C before further analysis. Total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) levels were measured enzymatically on a chemistry analyzer (Olympus AU5400; Chemical Ltd., Tokoy, Japan). Glucose levels were measured by a glucose oxidase method (Reagent kit; Diagnostic Chemicals Ltd., London, UK). High-sensitivity C-reactive protein (hs-CRP) was determined using an enzyme-linked immunosorbent assay kit (R&D systems, Minneapolis, MN, USA).

| Inflammatory cytokine measurements
Inflammatory cytokines including IL-1β, IL-8, and IL-11 were measured using the CBA Human Soluble Protein Detection Kit (BD Biosciences, San Jose, CA, USA) as previously described. 19 Briefly, blood samples and standards were incubated with capture beads for 1 hour at room temperature in the dark. The phycoerythrin detection reagent was then added and incubated for 2 hours at room temperature in the dark. The samples were washed and the bead pellets were resuspended in washing buffer. The resuspended samples were then run on a flow cytometer (FACS Canto II; BD Biosciences). Two thousand events in the gated bead population were collected, and 5-parameter data were saved for subsequent analysis using BD FCAP Array software. Serum concentrations were derived using the standard curve and expressed in pg/mL.

| Serum Metrnl measurement
Serum Metrnl levels were measured by using an enzyme-linked immunosorbent assay kit (R&D systems) according to the manufacturer's protocol. The intra-and interassay coefficients of variance were 2.59% and 3.55% respectively. The analytic sensitivity of the assays was 15.625 ng/mL.

| Statistical analysis
Statistical analyses were performed using PASW 18.0 (IBM SPSS Inc., Chicago, IL, USA). Normality of distribution was assessed using the Kolmogorov-Smirnov test. Data for age and LDL-C were normally distributed parameters and presented as the mean ± SD, and comparisons were analysed by Student's t test. Skewed data, including SBP, DBP, TC, TG, HDL-C, FBG, Cr, hs-CRP, IL-1β, IL-8, IL-11, Gensini score, and Metrnl, were expressed as median and quartile ranges, and comparisons were analysed by the Mann-Whitney U test. Pearson chi-squared test was used to compare qualitative variables represented as frequencies. The correlations between serum level of Metrnl and other variables were calculated using Spearman correlation coefficient and partial correlation coefficient adjusted for age, sex, smoking, alcohol intake, BMI, Cr, FBG, TC, TG, and LDL-C, as appropriate. Univariate analysis and multivariate logistic regression analysis were taken to determine the variables that independently contributed to the presence of CAD. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Linear regression analysis was used to test for trend of the changes of serum Metrnl concentrations across the severity of coronary angiography (normal to triple-vessel disease). Receiver operating characteristic (ROC) curve analysis was used to determine the optimum cut-off level of Metrnl best predicting CAD. To assess the performance of predictive model of Metrnl for CAD, we calculated the concordance (c) statistics for discriminative ability and goodness of fit statistics for calibration. All tests were two-sided and P < 0.05 was considered statistically significant.
To further explore the synergistic effect of Metrnl and classical risk factors on the risk of CAD, a 4 × 2 table approach was conducted to calculate OR, respective 95% CI, and two-tailed P-values.
It was assumed that subjects unexposed to the traditional risk factors and with a Metrnl concentration above the cut-off value have a certain background risk for CAD (defined as OR00, with the value assumed to be 1). OR01 refers to the relative risk for CAD among those unexposed to a risk factor with a Metrnl concentration below the cut-off value when compared with those with neither a Metrnl concentration below the cut-off nor exposure to risk factors. OR10 refers to the relative risk for CAD among those with a Metrnl concentration above the cut-off value but exposure to an environmental risk factor when compared with those with neither a Metrnl concentration below the cut-off value nor exposure to risk factors. OR11 is the ratio of the CAD risk among subjects exposed to risk factors with a Metrnl concentration below the cut-off value when compared with the CAD risk among unexposed subjects with a Metrnl concentration above the cut-off value. These ORs were used in the calculation of the synergy indexes (SIs), such that SI = (OR11-1)/ (OR10 + OR01-2) and SIM = OR11/(OR10 × OR01); the relative excess risk due to interaction, RERI = OR11-OR10-OR01 + 1; and the attributable proportion of the disease due to interaction, AP = RERI/OR11. Table 1 presents the characteristics of the study population. The 349 participants (mean age 65.1 ± 9.2 years) included 193 patients with CAD and 156 control subjects. There were no statistically significant differences between CAD patients and control subjects with respect to age, proportion of male, body mass index (BMI), rates of smoking and alcohol intake, rates of hypertension and hyperlipidaemia, TC, HDL-C, and creatine (Cr) levels. However, CAD patients, compared with control subjects, had higher rates of diabetes, higher levels of SBP and DBP (P < 0.01); higher levels of TG and LDL-C (P < 0.03); higher levels of FBG, hs-CRP, and IL-1β (P < 0.01); higher rates of taking antihypertension drugs, antihyperglycaemic drugs, and hypolipidaemic drugs (P < 0.001); and lower level of IL-11 (P = 0.005). In addition, the serum levels of Metrnl were lower in CAD patients when compared to the control subjects (median: 132.41 pg/mL vs 173.17 pg/mL, P < 0.001).      Metrnl levels were also decreased in patients with AMI ( Figure 4C) and negatively correlated with hs-CRP (r = −0.403; P = 0.001).

| Stratification analyses for the association of Serum Metrnl with the presence and severity of CAD
Previous studies revealed that Metrnl could antagonize insulin resistance and improve glucose tolerance in mice. 13 Figure S1). Moreover, we did not find significant difference of serum Metrnl between men and women in our study (median: 147.10 pg/mL vs 150.27 pg/mL, P = 0.366). Separating subjects into males and females also did not materially change the association of Metrnl with the presence and severity of CAD (Table 3 and Figure S2).

| Interaction between Metrnl and classical risk factors
As shown in Tables 4 and 5

| DISCUSSION
The main finding of our study is that serum Metrnl, a novel adipokine with protective effect on insulin resistance, was strongly associated with CAD in a Chinese population. Among 349 subjects enrolled in our study, patients with the lowest tertile of serum Metrnl level were associated with nearly 1.5-fold increase in the risk for CAD when compared to those with the highest tertile. Another interesting finding of the present study is that serum Metrnl level was also associated with the severity of CAD.  Tertile 2 vs Tertile 3  Tertile 1 vs Tertile 3 Crude OR (95% CI)

Pvalue
Adjusted OR (95% CI)  Another interesting finding of our study is that serum Metrnl concentrations were independently inversely correlated with metabolic parameters, including BMI, TC, and LDL-C. It has been long known that metabolic syndrome is associated with increased risk of CAD events. 21  macrophage activation associated with the suppression of inflammatory cytokines such as TNF-α, IFN-γ, and IL-1β, and the increase in anti-inflammatory genes such as IL-10 and TGF-β. 11 It has been well known that pro-inflammatory M1 macrophages participate in atherosclerosis initiation and progression while M2 macrophages are protective due to their anti-inflammatory properties, presumably stabilizing the plaque. 25 Considering that Metrnl is also highly expressed in heart and endothelial cells, 11,24 we speculate that decreased Metrnl might impair M2 polarization and potentiate inflammation of M1 macrophages, ultimately accelerating the development of atherosclerosis. Indeed, our study also found a negative correlation between serum Metrnl and pro-inflammatory markers including hs-CRP and IL-1β as well as a positive correlation between Metrnl and anti-inflammatory cytokine IL-11. However, further studies using animal or cellular models are needed to delineate the effects and mechanisms of Metrnl in the pathological process of CAD and atherosclerosis.

Pvalue
Our study should be interpreted within the context of its limitations. The cross-sectional design does not allow causal inference.
Secondly, the relatively small sample size may underpower the results of our study. Thirdly, the CAD patients of our study were actually from a referral population. This might cause the possibility of selection bias and confound the results. In addition, as we only investigated Metrnl levels in Chinese patients, our findings need to be confirmed in other ethnicities.

| CONCLUSIONS
In summary, our first case-control study shows significant associations of serum Metrnl with the presence and severity of CAD.
Although further prospective and interventional studies are necessary to investigate whether decreased Metrnl may participate in the development of atherosclerosis, our findings provide novel insights into the potential role of Metrnl in CAD.

CONFLI CT OF INTEREST
The authors declare that they have no competing interests.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE
Written informed consent was obtained from each participant and this study was approved by the Ethics Committee of the Affiliated Sir Run Run Hospital and Second Affiliated Hospital of Nanjing Medical University.

AUTHOR CONTRI BUTION
Wei Gao and Xiang Lu designed the study, interpreted the data, and